Michael

Excited to present Relay Block Merging (rBM), a joint exploration with the @gattacahq team- rBM increases block value by offsetting exclusive flow and timing games: → Higher revenue for proposers and builders. → Higher staking rewards. → Sustainable income for relays. It also improves censorship resistance by allowing relays to sort in transactions a builder may have filtered. Viewed another way, it improves throughput. Primer on why Relay Block Merging works: (1) Exclusive flow refers to transactions that are not publicly visible. This brings numerous advantages, like front running protection. However, exclusive flow can constrain block value. This is because individual builders may not see all transactions that could be combined into an optimal block. →Relay Block Merging offsets this by allowing relays to append transactions unavailable to the winning builder. →The merged block has to be more valuable than the best bid from the PBS auction at delivery time. →This strictly increases block value while leaving the execution guarantees of the winning builder intact. (2) Relay Timing Games refer to relays delaying the delivery of their best bid. As bids typically increases over time, this increases the relay's chance of winning the slot. → Relay block merging repurposes this idle time into a "merging stage" used to optimize block value. TL;DR → relay block merging is a way of increasing block value and censorship resistance today. → This benefits relays, builders, proposers, and stakers. Relay Inclusion Lists (rILs) are a complementary design: → rILs increase the censorship resistance of txs in the mempool. → rBM increases the censorship resistance of exclusive flow. More on rILs: x.com/gattacahq/stat…

@ericonomic @Kinetiq_xyz @Markets_xyz Congrats Eric. What a match!

I’m proud to share that I’ve joined @Kinetiq_xyz and @Markets_xyz. I’ll be focused across several verticals, including product, growth, and more. But one area I’m especially excited about is onchain analysis. My goal is simple: make sure every wallet worth watching on Hyperliquid is tracked in the Markets App. And through @MarketsAlpha, I’ll be sharing how I find them, why they matter and what they reveal about the market. Really excited for what’s coming. Put the k on the Hype.

Want a gentle intro to distributed computing for blockchain? I've been turning my notes "Consensus in 50 pages" into a short book called “A Quick Consensus”. Most of the way through, and what's there should already serve as a rigorous but accessible intro to the essentials. Covered: Tendermint, PBFT, HotStuff, Simplex, accountability, player reconfiguration, asynchronous SMR (and a bunch more). Still to add: erasure coding, DAG protocols, 2-round finality, recovery, and the Pipes model. Current draft (updated frequently): Lewis-Pye.com

@barnabemonnot @Uptodatenow The transaction pipeline will also play a large complementary role in scaling demand. A recent example are pAMMs giving better fills than CEXes. Predictable (multi-channel) inclusion is a big target. Censorship resistance for some, reduced adverse selection risk for others.

Finally read @Uptodatenow 's excellent overview of the strawmap! The strawmap is indeed mostly a supply-side endeavour. But I want to stress 2 of its 5 objectives that in my view have the potential to increase demand by improving service quality or offering net new features: 1. Fast L1 objectives (shorter slots, faster finality) directly improve UX, capital efficiency and onchain markets. See eg @benjaminion_xyz 's latest consensus.ethereum.foundation/blog/upgrading… or @jonah_b 's post from last year x.com/jonah_b/status… 2. Privacy at L1 also addresses what appears to be a strong user demand. Good to see steps in that direction with eg keyed nonces x.com/soispoke/statu… and other proposals like EIP-8182 x.com/dumbnamenumber…

@ThogardPvP The Swiss equivalent (GAP FEER) has 300 pages! For SMEs, granted.

@mostlyblocks It has to handle all of the edge cases, and there are quite a few

Simplicity should be a primary goal of any system that has to work for multiple parties. Example: The US Generally Accepted Accounting Principles total 25k pages. How can something like this be "generally accepted"?

@ThogardPvP Definitely. Grouping under “lack of information” as the top level bucket. Reminds me - some good reads on this in the monad neck of the woods

@mostlyblocks it's also a gnarly distributed systems problem / race condition

For optimal blockspace allocation, we need predictable inclusion. It should be ~predictable from: (1) The state of the chain. (2) The state the tx wants to access. (3) The tx's fee. The prediction can be complex, but it should be possible. Challenges are: - Deliberate exclusion, like censorship. - Accidental exclusion, due a lack of information. In short, censorship and information asymmetries like exclusive flow need to be addressed. We can address this by widening allocation from a single builder to multiple builders.

Concise argument worth reading. Particularly like the view of ETH as a commodity with functional utility. imv this utility is composed of: (1) Blockspace, which ETH buys. (2) Value, which ETH stores. High issuance for excess security is not conducive to either.

TL;DR Blockspace allocation is matching supply with demand. MEV refers to a laundry list of frictions than can be incidental to this process. MEV is not a precise term and that makes it difficult to work with. There has been very good early literature, but the term has largely outlived its usefulness.

Blockspace allocation is broader than MEV, which usually stands in for a loss term. The goal of blockspace allocation is to predictably match supply and demand. MEV has been shorthand for procedural hazards ranging from censorship to information leakage. Helpful concept esp. in the early days, but it's more of a subset of the problem. Likely also lost mindshare with searchers being supplanted by CEX/DEX firms.

@barnabemonnot @ethereumfndn @TimBeiko @ralexstokes All the best Barnabé. Enjoyed our occasional discussions, and curious what’s next.

Excited to announce this evolution of Protocol, the @ethereumfndn teams stewarding, researching and developing the Ethereum protocol. After our re-launch of Protocol in June last year, @TimBeiko, @ralexstokes and I are now passing the torch to our talented colleagues @corcoranwill @kevaundray and @fredrik0x. They are taking on the task of delivering on Scaling, UX and Hardness objectives, with the protocol strawmap in their pocket (strawmap.org). --- It is also time to announce that I made the decision to leave the Ethereum Foundation, my home for the past 6.5 years ❤️ I am so grateful for this opportunity I had, to work with amazing individuals, on the most impactful project there is. Looking back from when I started (here it is -> x.com/barnabemonnot/…), it has been a wild ride from early EIP-1559 work, to the Merge, to MEV markets, to staking, finality, interoperability and UX; and from my beginnings in the Robust Incentives Group to co-leading Protocol for the past year. Over this past year, our Protocol priorities, particularly our "Improve UX" work, shifted my attention to nearer-term questions. Throughout, I've been excited to take on a more product-centric view. Making Ethereum's unique features more available to users today is on my mind; so is participating in the plurality of ways that Ethereum gets built. I'd love to hear from friends old and new about what excites them at the moment, and share where I'm at. Please reach out!

Interesting proposal. This allows txs from a single sender to appear in a block in any order. Currently, same-sender txs have to appear in linear nonce order. A nice censorship resistance property when multiple builders contribute to blocks: Even if an earlier tx gets filtered, the remaining txs could still all find inclusion. Very relevant, as the block building pipeline is shifting in this direction. See e.g.: x.com/blockspaceforu…

@jgattermayer @d_gusakov @LidoFinance @EthPrague How can I get a copy @d_gusakov? Handy for friends & family..

what if... the real value was the friends we made on the way? thank you @d_gusakov and @LidoFinance for bringing me this rare crypto onboarding book at @EthPrague!

Frame Transactions TL;DR 1. Users define how their txs are approved and paid for. 2. Verification logic sits in a smart contract. Sample use cases are post-quantum signatures, gas abstraction, privacy.. Bottlenecks include verification and sequencing. There are responses in- and out-of-protocol. Verification is a bottleneck as frame transactions allow arbitrary approval logic. For this reason, the spec includes a 100k mempool gas limit. This is insufficient for complex applications, such as privacy protocols using groth16. To address this, @nero_eth proposes a higher ~400k limit for canonical smart contracts, which behave predictably. Sequential sequencing is a bottleneck when multiple users share a sender address, such as for privacy protocols. As nonces are currently sequential, this means that any delay affects all subsequent transactions. Keyed nonces proposed by @soispoke allow parallel transactions by allowing multiple nonce-domains. The out-of-protocol tx pipeline scales frame transactions: (1) Verification: Transactions of any size can be verified by builders. (2) Sequencing: (a) Keyed nonces remove sequential ordering requirements for single senders. (b) A single sender can find inclusion through multiple independent channels in a single slot. ->Doable through multi-party block construction (see @blockspaceforum Cannes workshop). V. important that multiple inclusion lanes exist out-of-protocol. Traders will continue to forego the mempool, but should still profit from predictable inclusion.

@Leozayaat Similar pattern in Switzerland surprisingly enough.

Ironically Singapore drivers are some of the most reckless drivers I’ve ever seen

A big plus of the MegaETH TGE is that it's taxed as capital gains, not income, due to the presale. Good structure. Proximity markets are interesting. Auctioning off the top seats is a understandable choice given token, but likely could get sig. more from direct deals. Clearly a lot of passion went into the UI, but not easy to find the "staking equivalent". Would be helpful as a header in rabbithole.

@d_gusakov @aiskivi @jdetychey @ethereum This assumes solo stakers would be more sensitive to a reduction in issuance than institutions. Why would this be the case? They are already foregoing profit by choosing to deploy themselves. Fixed costs are extremely minor.

@aiskivi @jdetychey @ethereum It is not as simple as a target percentage. What matters is who controls this percentage and how. 25% of ETH staked, with the majority controlled by CEXes and institutions, and we have chan that will censor users

If we cut @ethereum staking issuance, we will likely kill LSTs. Without LSTs, DeFi will shrink to the size of a penny. Without DeFi, @ethereum will lose its main value proposition. Without its value proposition, @ethereum will die. Do we really want to kill @ethereum?

@0xave Totally, great point. Platform, smart contract, and intermediary (eg issuer or bridge) risk.

@mostlyblocks Would say user demand also scales with trust in the underlying security of the smart contract (which is lowest it has been for a while)

DeFi lending yields are below treasuries because: A. Supply: On-chain cost of capital is lower than off-chain. This is likely due to a combo of taxes, compliance, and inertia. B. Demand: Limited due to limited user activity. As supply is abundant, the fundamental issue to address is user activity. This is the basis of real yields in DeFi and for stakers. User demand scales with three factors: (1) Good chain UX (2) Good applications (3) The crypto market (can't control) Chain UX is a prerequisite for good applications. If transactions are too expensive or too slow, there'll be no users. While the core protocol has to move at its own speed, UX can be much improved by the out-of-protocol block construction pipeline. For the chain, UX comes down to two things: (1) Cheap transactions. (2) Fast transactions. In other words: effective and fast blockspace allocation. We can improve blockspace allocation today by allowing multiple parties to contribute to block construction. Even early designs like relay block merging can increase blockspace utilization significantly. We can also allocate blockspace faster through sub-slots and preconfirmations. Many teams like Nuconstruct and ETHGas have done significant work on this. These services have been gated by the block construction supply chain, as multiple parties need to opt-in. Many teams are now aware that faster and collaborative blockspace allocation is the way to go, and a win<>win. We'll soon publish a concrete proposal for better blockspace allocation. It's a result of multiple workshops incl. in Cannes and Buenos Aires, and will be built in the open by multiple teams.

Not aware of any convincing arguments in favor of Ethereum's high issuance rate. There are "inaction" arguments that state that there's no clear alternative. This is a reasonable line early, but over time becomes lazy. The most frequent "pro" argument is that node operators will not be profitable if issuance is reduced. This restates the point, as staking beyond a target ratio should indeed not be profitable. Good reads: (1) @weboftrees 100-tweet magnum opus on minimum viable issuance: x.com/weboftrees/sta… (2) @pintail_xyz accessible and clear article on issuance today: x.com/pintail_xyz/st…